fluid-pr

作者: microsoft

在 Fluid Framework 儲存庫中建立拉取請求時使用。根據 Fluid Framework 慣例組合 PR 標題與內容,並向使用者提出建議…

npx skills add https://github.com/microsoft/fluidframework --skill fluid-pr
*CRITICAL* Do NOT run `gh pr create` until the user has explicitly confirmed via AskUserQuestion in step 5. Do not skip ahead.

CRITICAL Add the following steps to your task/todo list using your available task tooling (TaskCreate for Claude, TodoWrite for Copilot):

  1. Confirm you are NOT on main or any release branch. If you are, stop and tell the user: you cannot create a PR from a protected branch — they need to create or switch to a feature branch first.
  2. Verify that the origin remote does not point to microsoft/FluidFramework. If it does, stop and tell the user: pushing a branch directly to the main repo is not allowed — they should push to their fork instead. Pipeline changes on branches prefixed with test/ are an exception.
  3. Load the fluid-pr-guide skill NOW (via the Skill tool) before composing anything. It contains the title conventions, body template, and section guidance you need. Do NOT skip this step or rely on memory.
  4. Using the loaded fluid-pr-guide, compose the PR title and body following its conventions and template.
  5. Print the proposed title and body as text, then immediately use the AskUserQuestion tool to let the user choose what to do next. Use these exact options:
    • "Create PR" — Push the branch and open the pull request
    • "Create draft PR" — Push the branch and open a draft pull request
    • "Edit" — Revise the title or body before creating
    • "Cancel" — Don't create a PR
  6. If the user picks "Edit", apply their edits and re-present (go back to step 5). If "Create PR" or "Create draft PR", push and create accordingly. If "Cancel", stop.

Pushing and Creating the PR

Before pushing, verify that origin does not point to microsoft/FluidFramework. Run:

git remote get-url origin

If the URL contains microsoft/FluidFramework, stop unless the branch is prefixed with test/ and the conversation context or changed files indicate that the change needs Azure Pipelines testing. A user request for a "test branch" in that context is sufficient intent to use the exception. Otherwise, tell the user they should push to their fork instead and do not proceed.

Once the checks in steps 1–2 pass silently, compose the title and body, print them as text, then use the AskUserQuestion tool with the four options as described in step 5. This is the only point where the fluid-pr flow asks the user a question.

# Push branch (first time)
git push -u origin <feature-branch>

# Create PR (option 1)
gh pr create \
  --title "<title>" \
  --body "$(cat <<'EOF'
<body>
EOF
)"

# Create draft PR (option 2) — add the --draft flag
gh pr create \
  --draft \
  --title "<title>" \
  --body "$(cat <<'EOF'
<body>
EOF
)"

After creating the PR, output the PR URL so the user can navigate to it.

Updating an existing PR description

You do not have permissions to edit PRs on the upstream microsoft/FluidFramework repo via the API. If you need to update an existing PR's title or body, write the new content to a temp file and tell the user to copy-paste it into GitHub:

cat <<'EOF' > "$TMPDIR/pr-body.md"
<new body content>
EOF

Then tell the user: "I can't edit the PR directly — I've written the updated description to $TMPDIR/pr-body.md. Please copy-paste it into the PR on GitHub."

來自 microsoft 的更多技能

oss-growth
microsoft
開源增長駭客角色
agent-framework-azure-ai-py
microsoft
使用Microsoft Agent Framework Python SDK(agent-framework-azure-ai)构建Azure AI Foundry代理。适用于使用AzureAIAgentsProvider创建持久化代理、使用托管工具(代码解释器、文件搜索、网络搜索)、集成MCP服务器、管理对话线程或实现流式响应。涵盖函数工具、结构化输出和多工具代理。
development
airunway-aks-setup
microsoft
在AKS上設定AI Runway——從裸叢集到執行模型。涵蓋叢集驗證、控制器安裝、GPU評估、供應商設定及首次部署。時機:「設定AI Runway」、「上線AKS叢集」、「安裝AI Runway」、「airunway設定」、「部署模型至AKS」、「在AKS上進行GPU推論」、「在AKS上設定KAITO」、「在AKS上執行LLM」、「在AKS上使用vLLM」、「在AKS上設定模型服務」、「AI Runway控制器」。
devops
appinsights-instrumentation
microsoft
使用Azure Application Insights檢測Web應用程式的指南。提供遙測模式、SDK設定與組態參考。適用時機:如何檢測應用程式、App Insights SDK、遙測模式、什麼是App Insights、Application Insights指南、檢測範例、APM最佳實踐。
devops
applicationinsights-web-ts
microsoft
使用Application Insights JavaScript SDK(@microsoft/applicationinsights-web)為瀏覽器/Web應用程式進行檢測。適用於真實使用者監控(RUM)——頁面檢視、點擊、AJAX/fetch依賴、例外、自訂事件,以及與後端OpenTelemetry追蹤關聯的瀏覽器端GenAI代理追蹤。涵蓋SDK載入器指令碼與npm設定、框架擴充(React、React Native、Angular)、點擊分析、遙測初始化器,以及從瀏覽器發出的代理/工具/模型span的OTel GenAI語意慣例。
devops
azure-ai-anomalydetector-java
microsoft
使用適用於 Java 的 Azure AI 異常偵測器 SDK 建置異常偵測應用程式。在實作單變量/多變量異常偵測、時間序列分析或 AI 驅動監控時使用。
development
azure-ai-language-conversations-py
microsoft
使用 azure-ai-language-conversations Python SDK 實作對話語言理解(CLU)。當使用 ConversationAnalysisClient 分析對話意圖與實體、建置 NLP 功能,或將語言理解整合至應用程式時使用。
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python。用於機器學習工作區、作業、模型、資料集、計算資源與管線。 觸發詞:「azure-ai-ml」、「MLClient」、「workspace」、「model registry」、「training jobs」、「datasets」。
development